"About 5,000 lives could be saved each year in England if GPs follow new guidelines on cancer diagnosis, the health watchdog NICE says."—BBC News, 23 June 2015
"GPs are being given a step-by-step guide to spotting cancer following concerns that diagnosis delays are costing 10,000 lives a year."—Daily Mail, 23 June 2015
It's estimated that earlier diagnosis of cancer could save around 5,000 lives in England every year, according to the National Institute for Health and Care Excellence (NICE), which has today published new guidelines for GPs trying to interpret symptoms.
NICE pointed us to separate research which found earlier diagnosis as one of the things that could've helped prevent between 5,000 and 10,000 deaths a year in the late nineties, although so too could increasing the number of surgical interventions. It doesn't give an exact split on how many of these lives would be saved by earlier diagnosis alone, and we don't know how changes in diagnosis and in the patient population might have changed the figure since the nineties.
So it's wrong to say that delayed diagnosis alone causes 10,000 deaths on the basis of this evidence.
NICE says it got to its figure of 5,000 by halving the upper end of this estimate of avoidable deaths, as experts it's spoken to think about half the lives would be saved by earlier diagnosis.
The problem with that is that if they'd taken the lower end of the estimate, which was 5,000 avoidable deaths, then halving that would suggest 2,500 were due to delayed diagnosis.
It's easy to see why it might be difficult to put an exact number on this. Treating cancer is complicated, and for any given death it's difficult to know whether the outcome would have been different with earlier or different treatment.
There's a lot of uncertainty here and it might be better to talk about there being around between 2,500 and 5,000 lives which could be saved, based on the research NICE pointed us to and their assumptions.
It's reasonable to say that thousands of lives could be saved by earlier diagnosis but putting a precise figure on it risks creating a spurious impression of certainty.